MonteSUR for VECM

[macro_man]

All,

Are there any examples of doing near-VARs like that in the MonteSUR within the framework of a VECM? Thanks.

[TomDoan]

All,

Are there any examples of doing near-VARs like that in the MonteSUR within the framework of a VECM? Thanks.

With the cointegrating vector treated as known, there's nothing special for Monte Carlo integration about a VECM vs a VAR or any other type of systems regression. You just set up your system including the error correction series. The only tricky part is that, unlike a VECM that you can define with SYSTEM (with a full set of variables in each equation), there is no way to substitute back automatically for computing IRF's. Instead, you would have to include identities in the model used to do the IRF's that define the error correction series in terms of the original series.